Re: possible bug in 8.4

On 2008-12-19, at 13:07, Tom Lane wrote:

> Grzegorz Jaskiewicz <gj(at)pointblue(dot)com(dot)pl> writes:
>> Filter: (NOT (hashed subplan))
> ^^^^^^^^^^^^^^
>
> If 8.3 does that, and 8.4 doesn't, it's most likely because you are
> using different work_mem settings.

you're right, as always :)

My point is, why is planner choosing plan on 8.4 that's obviously more
expensive ? even without sort_mem (work_mem) set to higher value ?
Obviously the plan is quite expensive, so probably sorting it on disc
- would be still cheaper.
The example is quite unrealistic, but I am trying different simple
things to test planner differences between these two versions, and see
if there's any improvement/regression.
This time I will set both work_mem to same value :)